acta oecologica 33 (2008) 1–9 available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/actoec Original article Laurel forest recovery during 20 years in an abandoned firebreak in Tenerife, Canary Islands Jose´ Ramo´n Are´valoa,*, Marı´a Dolores Perazaa, Carlos A´ lvareza, Alfredo Bermu´deza, Juan Domingo Delgadob, Antonio Gallardoc, Jose´ Marı´a Ferna´ndez-Palaciosa aDepartamento de Ecologı´a, Facultad de Biologı´a, Universidad de La Laguna, La Laguna 38206, Tenerife, Spain bDepartamento de Fı´sica Ba´sica, Facultad de Fı´sica, Universidad de La Laguna, La Laguna, 38206, Spain cDepartamento de Ecologı´a, Universidad Pablo Olavide, Sevilla, Spain article info abstract Article history: This study assessed the recovery of the structure and species composition of a laurel forest Received 4 July 2006 in an abandoned firebreak in the Rural Park of Anaga, Tenerife (Canary Islands). We statis- Accepted 29 June 2007 tically compared values of species richness, density and biovolume between 23 plots in the Published online 16 August 2007 firebreak and six control plots in natural forest near the firebreak. We evaluated changes in species composition with detrended correspondence analysis (DCA) based on densities and Keywords: biovolume. Biovolume is increasing significantly along the successional gradient (from Biovolume 1990 to 2004) but remains less than the values in control plots. Stem densities were signif- Canary Islands icantly lower in control plots than in 2004 plots. Species richness was significantly higher DCA in control plots than in 2004 plots (although there were no differences in values obtained Densities between the first sampling period 6 years after abandonment, and the second sampling Disturbance 20 years after abandonment). Changes in species richness are significant, but all species Forest succession present in control plots are also found in the firebreak plots. DCA based on biovolume sig- nificantly discriminated control plots from firebreak plots in 1990 (for axis I). Results sug- gest recovery to a laurel forest is occurring, although more time will be required to reach control plot density and biovolume values. The low intensity of disturbance and a well- conserved forest adjacent to the firebreak favour the recovery of species inside the fire- break. We advise eliminating suckers from all small trees (leaving the bigger stems) to accelerate succession to a vegetation structure similar to that found surrounding the firebreak. ª 2007 Elsevier Masson SAS. All rights reserved. 1. Introduction The construction of roads and trails can result in habitat loss. Roads and trails can also act as corridors or habitat for species Today, habitat loss, primarily due to deforestation and land not found in the undisturbed forest matrix (Forman et al., use changes, is considered one of the most important threats 2002; Forman and Alexander, 1998; Spellerberg, 1998). Despite to species diversity in the world (Gurevitch and Padilla, 2004). some ecological similarities, forest firebreaks are clearly * Corresponding author. Fax: þ34 922 318 311. E-mail address: [email protected] (J. Ramo´ n Are´valo). 1146-609X/$ – see front matter ª 2007 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.actao.2007.06.005 2 acta oecologica 33 (2008) 1–9 different from other anthropogenic corridors (e.g. less impact height. Maximum canopy heights are found at valley bottoms, on soil, constant dispersion of propagules onto the firebreak). decreasing progressively towards the upper valley margins. Whereas the restoration of roads requires active and intense The laurel forest of Anaga contains a total of 19 tree species intervention (Holl et al., 2000; Switalski et al., 2004) due to (Santos, 1990). Dominant species include Laurus novocanarien- the impacts of traffic on soil characteristics, dispersion of ex- sis, Erica scoparia, Erica arborea, Ilex canariensis, Prunus lusitanica, otic species, elimination of a native seed bank, elimination of Myrica faya and Viburnum tinus. The dominance of a given spe- area due to pavement and other factors (Milberg and Lamont, cies depends on site conditions. For example, E. scoparia dom- 1995), the impacts of firebreaks on the forest are, in contrast, inates on forest ridges, L. novocanariensis in mesic zones and E. lower and restoration can be accomplished with minimal to arborea in more disturbed areas (Anon, 1973). Because the area no management following abandonment. of the firebreak is mainly in the southern aspect of the park, Today only 10% of the original laurel forest on Tenerife Is- dominant species are E. arborea, M. faya and I. canariensis. Fur- land remains unaltered (Santos, 1990). While natural distur- ther information on stand composition, structure and envi- bances and regeneration in laurel forests have been studied ronment in the study sites can be found in Are´valo (1998) (Are´valo and Ferna´ndez-Palacios, 1998; Are´valo et al., 1999; and Are´valo et al. (1999). Species nomenclature follows Are´valo and Ferna´ndez-Palacios, 2003), many of the long Izquierdo et al. (2001). term dynamics of laurel forests, like other mountain cloud The firebreak is located in El Moquinal. Its highest point is forests elsewhere, are largely undescribed (Hamilton et al., at 847 m asl. It runs 450 m north to south and is 10 m wide 1995). Regeneration after disturbance is one of the principal (Table 1). This firebreak was established in the 1960s when factors that structures plant communities (Terradas, 2001). fire control was very popular and was maintained until its The study of an abandoned firebreak offers a valuable oppor- abandonment in 1984 (due to the low number of wildfires in tunity to learn more about the long-term dynamics of a laurel the area). The original vegetation was a well conserved laurel forest since succession following this disturbance could be forest similar to the one surrounding the present firebreak, similar to succession following other disturbances such as and vegetation was cut to ground level. Maintenance was per- gap openings in the canopy. formed every 3–5 years by a team of 4–10 men. They cut all the The main aim of this study is to describe the recovery of the vegetation in a 10 m wide band just above ground level. structure of the vegetation (density and biovolume) and spe- cies composition after a 20-year period following the distur- 2.2. Sampling design bance (no machinery, but manual clearing of the firebreak at intervals of 3–5 years). Also, the firebreak is surrounded by We established 23 2 Â 8 m (16 m2) permanent plots in 1990 a well conserved forest matrix and asexual regeneration along the centre of the firebreak every 20 m from the highest through basal resprouting, which is common in this type of point, starting from the border (the last one was located at vegetation, can help this recovery. Application of our study re- 440 m). The long axis of each plot was parallel to the firebreak. sults could help guide the development of restoration plans, The plots were numbered from 1 (at the highest point) to 23. In land reclamation planning and the formulation of recommen- 2004 six control plots with similar dimensions were estab- dations for plantations. lished at 50 m intervals at a distance of 10–15 m from the firebreak. In each plot we counted and measured the height of all in- dividuals present by species. We also measured the stem di- 2. Materials and methods ameter along the larger axis and the perpendicular axis, in order to estimate the biovolume (Ferna´ndez-Palacios and de 2.1. Study site los Santos, 1996) of each individual plant in the plot. For shrubs and forbs, we measured the diameters from the top. The study was conducted in the Anaga Rural Park in the NE For trees, we estimated the two axes from the projected can- corner of Tenerife (Fig. 1), Canary Islands (28 190 N, 16 340 opy on the ground. The sampling was conducted from January W). The park encompasses a 7–8 million year old basaltic mas- to March in 1990 and in 2004, respectively 6 and 20 years after sif (Ancochea et al., 1990) covering about 130 km2. The park abandonment. represents 7% of Tenerife’s total area. Tenerife’s evergreen laurel forest has been extensively exploited since the arrival 2.3. Soil parameters of the Europeans in the 15th century (Parsons, 1981). The 10% of the forest that remains has been formally protected Nutrient variability could explain the existence of different since 1988 and is currently experiencing reduced human dis- patterns of regeneration between different areas. We exam- turbance and no reduction in area. No data are available about ined soil characteristics and nutrient availability across the the precise forest age, but aerial photographs from 1952 show plots to reveal possible effects on species distributions. We the forest in its current state, in terms of both extent and collected 1 kg soil samples in March 2004 from four randomly physiognomy. In the 1940s, there was still some illegal, selected locations on the border of each plot by excavating to small-scale forest exploitation. Due to its present protection a depth of 15 cm. The four samples were combined to form status and public use, the illegal forest exploitation that did a composite sample for the analyses. We sieved (>2 mm) the occur was of limited extent (small scale). air-dried soil and aerobically stored each sample at room tem- The laurel forest is distributed between 600 and 900 m alti- perature. Litter covering the plot (not decomposed) was col- tude and, depending on the slope, the canopy is 10–20 m in lected and weighed in situ and re-distributed over the plot acta oecologica 33 (2008) 1–9 3 Fig. 1 – Actual (bold) and potential (plain) distribution of the laurel forest on Tenerife.